While live bacteria commonly cause infection, the relative role of bacterial toxins versus live bacteria to produce morbidity and mortality is not known, If bacterial toxins are important factors, therapies directed at these toxins may be required. Even if no bacteria remain alive or grow, these preformed toxins may be important in determining morbidity and mortality. Therapies directed at these toxins may be important additions to antibiotic therapy. Using two E. coli nonvirulent E. coli (O86:H8) and a virulent E. coli (06:41 :K2) in the canine model, we designed a study to determine whether organism virulence factors or bacterial toxins were more important in producing septic shock. Measures of hemodynamic shock, blood cultures, endotoxin levels and survival were done serially with both organisms live, both organisms killed, and with purified endotoxin from these organisms. Approximately 4 man years have gone into this study. Preliminary results from these studies have been presented at the American Federation of Clinical Research and the Society of Critical Care Medicine. These data suggested virulence factors are much more important than performed endotoxins on development of endotoxemia. To further address this issue we are collaborating with Tom Russo of LCI, NIAID and more recently the University of Buffalo to examine the role of endotoxin as a virulence factor within the same strain of E. coli. We are using a gene disruption technique to produce in an E. coli strain with a deficiency in endotoxin o type side chains (another virulence factor). This will further explore the relative role of endotoxin in septic shock. Much work is presently being done to target this pathogen in human septic shock. There has been much interest in these data as determination of the factors that produce morbidity in infection will guide trials for future therapies of children and adults with septic shock.